

COPYRIGHT © 2002 BY THE JOURNAL OF BONE AND SURGERY, INCORPORATED Isolated Gastrocnemius Tightness BY CHRISTOPHER W. DIGIOVANNI, MD, RODERICK KUO, MD, NIRMAL TEJWANI, MD, ROBERT PRICE, MSME, SIGVARD T. HANSEN JR., MD, JOSEPH CZIERNECKI, MD, AND BRUCE J. SANGEORZAN, MD Investigation performed at the Department of Orthopaedics, Harborview Medical Center, and the Seattle Veterans Affairs Medical Center, Seattle, Washington

Background: of the gastrocnemius-soleus complex has well-documented deleterious effects on lower-limb function in spastic or neurologically impaired individuals. There is scarce literature, however, on the existence of isolated gastrocnemius contracture or its impact in otherwise normal patients. We hypothesized that an inability to dorsiflex the due to equinus contracture leads to increased pain in the forefoot and/or midfoot and therefore a population with such pain will have less maximum ankle dorsiflexion than controls. We further postulated that the difference would be present whether the was extended or flexed. Methods: This investigation was a prospective comparison of maximal ankle dorsiflexion, as a proxy for gas- trocnemius tension, in response to a load applied to the undersurface of the in two healthy age, weight, and sex-matched groups. The patient group comprised thirty-four consecutive patients with a diagnosis of meta- tarsalgia or related midfoot and/or forefoot symptoms. The control group consisted of thirty-four individuals without foot or ankle symptoms. The participants were clinically examined for gastrocnemius and soleus con- tracture and were subsequently assessed for tightness with use of a specially designed electrogoniometer. Measurements were made both with the knee extended (the gastrocnemius under tension) and with the knee flexed (the gastrocnemius relaxed). Results: With the knee fully extended, the average maximal ankle dorsiflexion was 4.5° in the patient group and 13.1° in the control group (p < 0.001). With the knee flexed 90°, the average was 17.9° in the patient group and 22.3° in the control population (p = 0.09). When gastrocnemius contracture was defined as dorsi- flexion of ≤5° during knee extension, it was identified in 65% of the patients compared with 24% of the control population. However, when gastrocnemius contracture was defined as dorsiflexion of ≤10°, it was present in 88% and 44%, respectively. When gastrocnemius-soleus contracture was defined as dorsiflexion of ≤10° with the knee in 90° of flexion, it was identified in 29% of the patient group and 15% of the control group. Conclusions: On the average, patients with forefoot and/or midfoot symptoms had less maximum ankle dorsi- flexion with the knee extended than did a control population without foot or ankle symptoms. When the knee was flexed 90° to relax the gastrocnemius, this difference was no longer present. Clinical Relevance: These findings support the existence of isolated gastrocnemius contracture in the develop- ment of forefoot and/or midfoot pathology in otherwise healthy people. These data may have implications for preventative and therapeutic care of patients with chronic foot problems.

ightness of the gastrocnemius-soleus complex has long first described in the early 1900s by Silfverskiold17 and by been documented in spastic and neurologically im- Vulpius and Stoffel 18, respectively, albeit in spastic patients19-21. paired individuals1-3. Since the first description of To date, even the definition of and method of examining for T 4 tendo Achillis lengthening in the early 1800s by Delpech , re- equinus contracture remain controversial, with the maximal lease or attenuation of the superficial posterior compartment ankle dorsiflexion values used as a measure of equinus defor- of the leg has been performed in many ways to relieve equinus mity ranging from 0° to 25° and the examination performed contracture and to improve gait and muscle balance across the with the knee in varying amounts of flexion or extension22-28. foot and ankle5-16. Very little attention, however, has been paid As a result of such confusion, the reliability of clinical exami- to the cumulative pathological effects or even the existence of nation for correctly identifying contracture also has never a more subtle equinus contracture of the gastrocnemius that been determined. can be found in the “normal,” otherwise unaffected popula- Except for a few still controversial examples of plantar tion. This is surprising considering that isolated gastrocne- , forefoot ulceration in diabetics, or progressive hallux mius tightness and its treatment with surgical recession were valgus or flatfoot, the relationship between tightness of the su-

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014 

THE JOURNAL OF BONE & JOINT SURGERY · JBJS.ORG ISOLATED GASTROCNEMIUS TIGHTNESS VOLUME 84-A · NUMBER 6 · JUNE 2002

perficial posterior compartment and progressive pathological the study and, if they were, to sign an informed-consent form. changes in the foot in non-spastic individuals has been over- looked entirely by the orthopaedic community26,29-32. In contra- Clinical and Radiographic Examination distinction, more attention has been paid to this phenomenon A brief medical and surgical history was obtained from each in the podiatric literature over the past three decades, although individual to be studied, and a directed most of those reports are observational or anecdotal10,23,28,33-38. of both feet and was performed. The patient’s height We know of no paper specifically addressing the prevalence of and weight, any previous treatments, and all medications isolated gastrocnemius contracture or its long-term effects in were recorded. Any history of smoking, diabetes, or ligamen- otherwise normal, healthy people. tous laxity (defined as elbow hyperextension) was also noted. Our experience has suggested that the gastrocnemius Examination was performed to identify any physical or struc- muscle in particular is the predominant deforming force in tural foot or ankle abnormalities, with attention paid to the people with structural breakdown or chronic pathological chief symptom. A subjective clinical assessment of the pres- changes related to the foot and ankle. We suspect that con- ence or absence of equinus was then made by the examiner. tracture of this muscle is not only common but often partially Gastrocnemius contracture was simply recorded as “present” responsible for many of the chronic forefoot and midfoot or “absent” by the examiner on the basis of the Silfverskiold symptoms identified in non-neurologically impaired patients. test17. Weight-bearing anteroposterior, lateral, and oblique The purpose of our study was to identify the prevalence of radiographs of the foot and/or ankle were not routinely made equinus contracture in a group of individuals with and with- for the purposes of this study, but any that had been made out foot symptoms, with a specific focus on the presence of incidentally during a patient’s clinic visit were evaluated. No isolated gastrocnemius tightness and our ability to accurately diagnose it. We hypothesized that an inability to dorsiflex the ankle as a result of equinus contracture leads to increased pain in the forefoot and/or midfoot and therefore a population with such pain will have less maximal ankle dorsiflexion than controls. We further postulated that the difference would be present whether the knee was extended or flexed.

Materials and Methods Patient Enrollment ubsequent to obtaining Institutional Review Board ap- Sproval, we prospectively identified sixty-eight subjects for this investigation between April 1, 1999, and July 31, 1999. The patient population comprised thirty-four consecutive pa- tients who presented to our University and Veteran’s Adminis- tration foot and ankle clinics because of isolated forefoot or midfoot pain of any duration. The exclusion criteria for the approximately 1000 patients screened for this group included any history of neurological disease including neuroma, any systemic disease potentially affecting the foot or ankle such as the vasculitides or , any confounding pre- existing foot or ankle surgery or trauma, any osseous block to ankle dorsiflexion, or any irreducible foot deformity (i.e., one making it impossible to obtain a neutrally aligned foot by re- ducing the talonavicular joint) precluding proper testing and evaluation of ankle equinus. Individuals with symptoms re- lated solely to the hindfoot or ankle, those with severe hind- foot deformity, and those with midfoot pain suspected to be the result of ankle or hindfoot pathology were also excluded. Fig. 1 The control population consisted of thirty-four randomly The equinometer comprises an electrogoniometer (A) attached to the lat- identified spouses of patients or hospital personnel from mul- eral aspect of the leg by means of a four-bar linkage (B), which is con- tiple services who had no foot or ankle symptoms. Approxi- nected to a rigid foot-plate attached underneath the foot with a force mately seventy-five people were interviewed for inclusion in transducer (C) positioned beneath the second metatarsal head. The an- the control group before a sufficient number who met the ex- kle acts as the center of rotation for an upwardly applied force to the clusion criteria described above could be enrolled. All people plantar surface of the foot. Resultant maximal ankle dorsiflexion is mea- fulfilling these criteria were then approached by an indepen- sured by the computer (D), which accrues data only when the upward dent provider to ask them if they were willing to participate in force on the foot is appropriate for a predetermined torque of 10 N-m.

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014 

THE JOURNAL OF BONE & JOINT SURGERY · JBJS.ORG ISOLATED GASTROCNEMIUS TIGHTNESS VOLUME 84-A · NUMBER 6 · JUNE 2002

radiographs had been made for any person in the control ible anatomic landmark and has an axis that closely approxi- group. mates the center of rotation of the ankle joint. The foot-plate and accompanying force transducer were then attached to the Equinometer Testing undersurface of the neutrally aligned foot (Fig. 3). The talo- Maximal ankle dorsiflexion was measured with an equino- navicular joint is held reduced by the examiner, thus reduc- meter affixed to the subject’s leg (Fig. 1)39,40. Data from this ing the hindfoot and locking the midfoot. Midtarsal motion electrogoniometric device were processed by a Macintosh G3 is prevented because the cup is attached to a rigid steel series computer software package (Apple Computer, Cuper- bar the full course of the foot. With the foot held in tino, California), which depicts ankle position in the sagittal neutral in the device, ankle dorsiflexion could be measured plane as a result of an upwardly directed load applied to the after the 10-N-m load was applied to the undersurface of the undersurface of the foot. The site of torque application is di- foot. Because various data suggest that the gastrocnemius rectly under the second metatarsal head, the approximate cen- muscle is under no tension with the knee flexed ≥25°, ankle ter of force transmission in the human foot. The instrument dorsiflexion was recorded with the knee in full extension and shows a reading of 0° when attached to a neutral, plantigrade in 90° of flexion42. Three recordings of maximal resultant an- foot, with subsequent positive change indicating relative dor- kle dorsiflexion were obtained for each knee position. Sub- siflexion and negative change indicating plantar flexion. A jects were randomized (alternated) with regard to whether predetermined, constant torque of 10 N-m was arbitrarily se- testing was started with the knee in extension or flexion. lected for use in this experiment. This value represents the These latter two steps were taken to isolate independent con- average pressure placed on the foot by two of our orthopae- tributions of both the gastrocnemius and the to dists and one rehabilitation medicine specialist while testing any existent contracture and to negate any effect for equinus contracture in an office setting. We had previously that repetitive testing might have on the tissues. Twenty-eight validated the equinometer in our laboratory for such use. The people in the patient group and thirty-three people in the instrument is accurate to within 1.31°, with a documented control group were tested with the equinometer on the left average daily variation in individual subjects of <0.5° (0.45° ± side, and twelve and fifteen, respectively, were tested with it 0.43° [standard deviation])41. Both a right-sided and a left- on the right side. There were six bilateral examinations in the sided equinometer were used, with recalibration to a plastic patient group and fourteen in the control population. All ex- foot and ankle model every few weeks to ensure reproducibil- aminations and equinus testing were performed by a single ity and accuracy. examiner. Each subject to be tested was seated on an examination The data accrued from this testing were used to compare table with the bare foot and leg exposed. Care was taken to the subjective determinations (by the clinician) and the objec- ensure that the subjects were relaxed. The fibula and the sec- tive determinations (with the equinometer) of the presence or ond metatarsal head were then identified by the examiner. A absence of equinus. Since absolute criteria for identifying gas- measurement from the tip of the fibula to the center of the trocnemius contracture either subjectively or objectively have second metatarsal head (i.e., the moment arm) for each per- yet to be established and we are still in the early phase of study- son was recorded in the computer (Fig. 2). The device was ing this problem, we used two cutoff determinants, ≤5° and carefully positioned alongside the lateral aspect of the leg in ≤10° of maximal ankle dorsiflexion after force application, as line with the fibula. This reference serves as a very reproduc- potential definitions of gastrocnemius contracture. The four

Fig. 2 Moment-arm measurement of the distance (M) between the distal tip of the lateral malle- olus and the center of the second metatarsal head (as extrapolated along the lateral col- umn of the foot) standardized a constant torque of 10 N-m for each individual. This measurement (M), in the equation torque = force × moment (M), allows the computer to provide feedback on appropriate upward pres- sure (force) to keep torque constant.

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014 

THE JOURNAL OF BONE & JOINT SURGERY · JBJS.ORG ISOLATED GASTROCNEMIUS TIGHTNESS VOLUME 84-A · NUMBER 6 · JUNE 2002

analyses included (1) clinical determination versus equinome- TABLE I Primary Diagnoses in the Patient Group ter measurement of ≤5° with the knee flexed, (2) clinical deter- mination versus equinometer measurement of ≤5° with the No. of knee extended, (3) clinical determination versus equinometer Diagnosis cases ≤ measurement of 10° with the knee flexed, and (4) clinical de- 7 termination versus equinometer measurement of ≤10° with the knee extended. Thus, for each individual, the presence or ab- Morton foot 13 sence of gastrocnemius contracture was decided subjectively Idiopathic Lisfranc arthrosis 1 on the basis of clinical opinion and objectively on the basis of Posterior tibial tendon insufficiency 9 two different criteria: maximal ankle dorsiflexion of ≤5° or of ≤10° with the knee fully extended. The examiner was con- 3 sidered to have made a correct diagnosis when the clinical Stress fracture (5th metatarsal) 1 decision (present or absent) correlated with the objective de- termination as defined by the above criteria in each scenario. Gastrocnemius-soleus contracture was arbitrarily defined as 120 to 260 lb [54.4 to 117.9 kg]) in the control group. The pri- present when the individual exhibited ≤10° of maximal ankle mary diagnoses, determined on the basis of the patient’s his- dorsiflexion in 90° of knee flexion as determined objectively tory, the results of physical examination, and radiographs, in with the equinometer. the patient group were classified into six categories (Table I). Statistical analysis was performed between matched Although only seven of the thirty-four patients were noted to groups with the use of the chi-square test and the Student have isolated metatarsalgia, nineteen (56%) had some compo- paired t test. The p value was set at 0.05 for determination of nent of metatarsalgia as part of the chief symptom. We defined significance. Study power was calculated for the resultant max- a Morton foot as a constellation of symptoms commonly imal ankle dorsiflexion values. identified in our clinics: hallux valgus, extensor recruitment, medial column , and a long second metatarsal Results with overload. Patients with insufficiency of the posterior tib- ixty-eight people were included in this investigation. The ial tendon were found to have either stage-one or stage-two Saverage age of the patient group was forty-seven years disease43. We describe this deformity as dorsolateral peritalar (range, twenty-one to seventy-six years) compared with forty- subluxation, and we excluded any patient with stage-three dis- five years (range, twenty-eight to sixty-three years) for the ease because feet with irreducible deformity cannot be prop- controls. There were eighteen men and sixteen women in the erly tested for equinus contracture. We noted one stress patient population compared with sixteen men and eighteen fracture, at the base of the fifth metatarsal without antecedent women in the control group. The average weight of the in- trauma. In the control group, all subjects were found clinically dividuals was 183 lb [83 kg] (range, 104 to 345 lb [47.2 to to have structurally sound feet. 156.5 kg]) in the patient group and 171 lb [77.6 kg] (range, A significant difference was found between the patient

Fig. 3 Achievement of a neutrally aligned foot in the jig was an extremely important step in testing tightness of only the gastrocnemius-soleus complex.

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014 

THE JOURNAL OF BONE & JOINT SURGERY · JBJS.ORG ISOLATED GASTROCNEMIUS TIGHTNESS VOLUME 84-A · NUMBER 6 · JUNE 2002

TABLE II Comparison of Average Maximal Ankle Dorsiflexion Between the Two Groups with the Knee in Either Extension or Flexion

Maximal Ankle Dorsiflexion* Knee in Knee in 0° of Flexion 90° of Flexion

Patient group 4.5° ± 4.5° (–1.1° to 14.3°) 17.9 ± 9.0° (6.1° to 37.0°) Control group 13.1° ± 8.2° (1.1° to 30.7°) 22.3° ± 10.9° (7.6° to 51.1°) P value <0.001 0.09 Power ratio 0.99 0.40

*The values are given as the average and the standard deviation, with the range in parentheses.

TABLE III Comparison, Between Groups, of Number of Subjects in Whom Gastrocnemius Contracture, Identified with the Equinometer, Was Diagnosed Clinically

Gastrocnemius Contracture* Correct Diagnosis*† Group ≤5° ≤10° ≤5° ≤10°

Patient 65% (22/34) 88% (30/34) 76% (26/34) 88% (30/34) Control 24% (8/34) 44% (15/34) 94% (32/34) 79% (27/34)

*≤5° and ≤10°, which refer to the maximal ankle dorsiflexion, indicate the two possible definitions of contracture. †The correct diagnosis re- fers to the correct identification of gastrocnemius contracture as either present or absent.

and control populations with regard to the average maximum three in the control group. Only one individual in the patient ankle dorsiflexion with the knee in the extended position (the group had compared with five in the con- gastrocnemius under tension) as opposed to the flexed position trol group. (the gastrocnemius relaxed) (Table II). The difference between Regardless of whether we defined gastrocnemius con- the ankle dorsiflexion range with the knee flexed and that with tracture as ≤5° or ≤10° of ankle dorsiflexion during full knee the knee extended was 11.5° ± 6.0° in the patient group com- extension, there seemed to be at least a twofold increase in the pared with 8.4 ± 4.8° in the control group (p = 0.02). rate of equinus in the patient group compared with that in the No significant difference was identified, during knee ex- control population and an approximately 85% rate of correct tension (p = 0.68) or flexion (p = 0.67), between the results for clinical diagnosis of this entity (Table III). Similarly, the num- the right and left sides of patients who had a bilateral exami- ber of individuals in the patient group exhibiting a so-called nation in either group. There was also no difference in either contracture (defined as combined equinus of group between the readings for the first and last (third) trials the gastrocnemius and soleus muscles) was nearly double that (p = 0.6 to 0.9) or between male and female subjects (p > in the control group (Table IV). 0.05). The average maximal ankle dorsiflexion in knee exten- sion was found to be independent of the initial starting posi- Discussion tion of the knee (either flexion or extension) (p > 0.05). or two centuries, much attention has been paid to equi- Similarly, there was no significant difference between the re- Fnus contracture of the gastrocnemius-soleus complex in sults of the whole patient group and those in the subpopula- patients with neurologic or spastic imbalance3,5-16. Associated tion of patients who presented with only metatarsalgia (p > long-term sequelae of ankle equinus, such as gait disturbance 0.05). When this subpopulation was compared with the con- and structural breakdown of the foot and ankle, have also trol group, however, the difference in average maximal ankle been well documented in these patients2,17,20,22,25,26,28,31,35,44,45. Such dorsiflexion was found to be significant (p = 0.001). equinus positioning of the foot has frequently been called an Despite the small numbers of individuals, there were a Achilles contracture. However, this is a misnomer because the few evident differences between groups with regard to some of majority of the perceived stiffness or stretch occurs within the the medical parameters that we measured. Five individuals in muscle bellies themselves, not in the tendon; the tendon can the patient group were found to have diabetes mellitus as op- be responsible for only about 3% to 5% of this change in po- posed to none in the control group. Seven individuals in the sition. Hence, although we often speak of “Achilles” tightness patient group had a history of smoking compared with only and address the Achilles tendon directly to treat this disorder,

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014 

THE JOURNAL OF BONE & JOINT SURGERY · JBJS.ORG ISOLATED GASTROCNEMIUS TIGHTNESS VOLUME 84-A · NUMBER 6 · JUNE 2002

TABLE IV Comparison, Between Groups, of Number of Subjects with Either No Evidence of Gastrocnemius Contracture or Evidence of Gastrocnemius-Soleus Contracture

No Gastrocnemius Contracture* Gastrocnemius-Soleus Group ≤5° ≤10° Contracture† Patient 35% (12/34) 12% (4/34) 29% (10/34) Control 76% (26/34) 56% (19/34) 15% (5/34)

*≤5° and ≤10°, which refer to the maximal ankle dorsiflexion, indicate the two possible definitions of contracture. †Gastrocnemius-soleus contracture was defined as ≤10° of maximal ankle dorsiflexion in 90° of knee flexion.

the root of the problem is actually at the level of the gastroc- 90°. The modicum of information on this seemingly underap- nemius and soleus muscles. Since 1816, when the first tendo preciated phenomenon is anecdotal and is mostly in the podi- Achillis lengthening was described in France, multiple meth- atric literature23,25,28,35. Since the early 1970s, podiatrists have ods of managing this problem have been reported4,10-16,18,20. sought to define and diagnose gastrocnemius and combined Improved gait, decreased muscular imbalance, and slowed equinus deformity, with some disagreement10,24,26-28. The maxi- progression of foot deformity have often been noted after mal ankle dorsiflexion values used as a measure of equinus these therapies in affected patients2,6,7,9,16,19,27,28,39. deformity have ranged from 0° to 25°, and various knee posi- Many foot deformities, such as hallux valgus, plantar ul- tions have been used to calculate these values22-28. A number of ceration, stress fracture, peritalar subluxation (acquired flat- recent reports have corroborated evidence linking equinus re- foot), and metatarsalgia, seem to have a multifactorial origin lease with decreased foot symptoms and breakdown as a result in neurologically “normal” people. Few orthopaedic papers, of improved ankle dorsiflexion, particularly in diabetics44,49-51. however, have discussed the role of a tight superficial posterior Subotnick28, Downey and Banks23, and Hill35 independently compartment in these people; furthermore, to our knowledge, emphasized improvement in gait mechanics and foot function none have really addressed the existence or role of an isolated after the effects of gastrocnemius or soleus tightness had been gastrocnemius contracture26,29,30,32,46. Most of these few reports alleviated in non-spastic individuals; Subotnick stated that deal with the relationship between a tight heel cord and flat- “gastrocnemius or soleus equinus is the greatest symptom pro- foot. Lin et al.47 also reiterated the importance of a tight heel ducer in the human foot.”28 In fact, two podiatrists described cord in diabetics, noting a markedly improved rate of healing the first (to our knowledge) isolated gastrocnemius lengthen- of plantar ulcers treated with a combination of a total-contact ing for treatment of a non-neurological deformity in 197452. cast and heel-cord lengthening as opposed to a cast alone; pre- Although these papers would be heavily criticized on a scien- sumably the improvement was due to decreased foot pres- tific basis by today’s standards, they linked equinus deformity sures. Brodersen et al.22 followed more than 750 children over with such foot conditions as , plantar fascii- the course of their education in the Danish school system and tis, metatarsalgia, peroneal tendinitis, Morton neuroma, and found that, in the two oldest classes, students with a high arch heloma durum and provided some of our only insight into the had a 39% prevalence of a tight superficial posterior compart- role of the gastrocnemius under normal conditions. Further- ment compared with prevalences of 56% and 77%, respec- more, Grady and Saxena24 challenged our ability to stretch a tively, in those with a normal arch and those with a depressed tight gastrocnemius-soleus complex, noting only a few degrees arch (flatfoot). They recommended stretching of the gastroc- of change in muscle flexibility after twenty-six weeks of a com- nemius-soleus complex in all school-age children. Some pa- monly prescribed home stretching regimen. This technique le- pers, such as that by Sinacore48, have described an exhaustive vers the foot between a dominant muscle group on one side list of risk factors associated with foot problems but have and the unyielding ground on the other. If the muscle does not never mentioned a tight superficial posterior compartment. actually stretch, such a combination could lead eventually to While equinus contracture is well recognized as the foot breakdown. inability to dorsiflex through the tibiotalar joint, to our knowl- We believe that we were the first to attempt to prospec- edge there is still no good study providing a numerical defini- tively identify (1) the prevalence of isolated gastrocnemius tion of gastrocnemius tightness or describing its prevalence, contracture in neurologically normal people, (2) its potential natural history, or treatment or the clinical ability to diagnose relationship with forefoot and/or midfoot pain, and (3) our the condition in humans. This is probably due to the subtleties ability to correctly diagnose it during physical examination. of midfoot motion and hence the clinician’s difficulty in accu- On the basis of the data obtained from our research, we be- rately performing an evaluation for equinus contracture. Cor- lieve that the inability to dorsiflex the ankle due to gastrocne- rect examination requires locking of the talonavicular joint, mius tightness represents a real entity in “normal” patients, and hence neutralization of the hindfoot, and subsequent eval- and, although we have not proved causality, we demonstrated uation of the change between the ankle dorsiflexion possible its significant association with increased pain in the forefoot with the knee flexed 0° and that possible with the knee flexed and/or midfoot. Patients with such pain had an almost three-

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014 

THE JOURNAL OF BONE & JOINT SURGERY · JBJS.ORG ISOLATED GASTROCNEMIUS TIGHTNESS VOLUME 84-A · NUMBER 6 · JUNE 2002

fold higher prevalence of isolated gastrocnemius contracture during knee extension; it increased to 88% when the defini- (when defined as ≤5° of dorsiflexion during knee extension) in tion was ≤10°. Although gait analysis data suggest that the av- comparison with a control population without foot symp- erage individual relies maximally on about 10° of ankle toms. They also had twice the prevalence of combined gas- dorsiflexion during the late-stance phase of normal walking53, trocnemius-soleus contracture compared with controls. When the gastrocnemius equinus remains ill-defined. We have se- evaluated with the knee in extension, these patients had only lected ≤5° of maximal ankle dorsiflexion with the knee in full one-third (4.5°) of the average maximal ankle dorsiflexion extension as our definition because it allowed us to diagnose demonstrated by the asymptomatic group (13.1°). However, the problem in those who were at risk (symptomatic patients) no significant difference in dorsiflexion was identified be- with fairly good reproducibility (76%) and, more importantly, tween the two groups when the knee was flexed and the gas- we were able to reliably avoid (in 94% of the cases) unneces- trocnemius was relaxed. These data support the concept that sary treatment of those who were not at risk (asymptomatic chronic mechanical overload of the foot and thus inflamma- people). This definition is complemented by our chosen defi- tion or wear and tear can result from equinus contracture nition of combined gastrocnemius and soleus contracture as (Fig. 4). Because it crosses the knee, ankle, and , ≤10° of maximal dorsiflexion with the knee in 90° of flexion the seems the most prone to being tight (i.e., regardless of knee position). in the leg and would therefore have the greatest effect during Interpretation of these findings is limited by the short- upright alignment. comings confounding our study. Generalization of our data This study also indicates that isolated gastrocnemius would have been more accurate with a larger number of pa- tightness in otherwise normal, healthy patients seems to be tients in both groups and with the inclusion of multiple subtle and that our ability to diagnose it clinically is, at best, blinded examiners. Also, our measurements relied on ana- only good. Our ability to make this diagnosis correctly in the tomic landmarks and thus were not exact, and weight-bearing patient group was only 76% when we used a definition of ≤5° radiographs were not compared between the two groups. In addition, electromyographic data may have permitted an as- sessment of the impact on patients of dorsiflexion end points, and application of 10 N-m of force on our equinometer was considered the so-called gold standard against which the clini- cal diagnosis was compared. We also included diabetic pa- tients, who could be considered to have a form of neurological impairment, and, while we extrapolated data from ankle ex- cursion with varying knee positions as an approximation of gastrocnemius tightness, it is not an exact measurement of ac- tual muscle tension. In summary, we believe that this report provides the first evidence of isolated gastrocnemius contracture and its poten- tial relationship with foot symptoms in otherwise healthy, non- neurologically impaired individuals. It demonstrated that (1) patients with metatarsalgia or related forefoot and/or midfoot symptoms had less average maximal ankle dorsiflexion with the knee in extension than did a control population without foot symptoms, and (2) when the knee was flexed to 90° to re- lax the gastrocnemius, this difference was no longer present. The literature remains confusing regarding the defini- tion of isolated gastrocnemius or combined gastrocnemius- soleus tightness. On the basis of the data obtained in the present investigation, we offer the following standardization of terms to describe the inability to dorsiflex through the tibiotalar joint: gastrocnemius equinus represents maximal ankle dorsiflexion of ≤5° with the knee in full extension, and Achilles tightness represents maximal ankle dorsiflexion of ≤10° with the knee in 90° of flexion. To our knowledge, this Fig. 4 paper is also the first to define an approximate prevalence of Graphically depicted load patterns during gait. This illustration de- both gastrocnemius and gastrocnemius-soleus contracture in monstrates the anticipated variation in forefoot and midfoot loading the general population. Depending on the chosen definition during late-stance-phase gait between “normal” subjects (i.e., those of contracture, roughly three-quarters of the patients with a with normal gastrocnemius length) and patients with an abnormal symptomatic foot and ankle in this study had isolated gastroc- (i.e., contracted) gastrocnemius length. nemius contracture and roughly one-third had a combined

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014 

THE JOURNAL OF BONE & JOINT SURGERY · JBJS.ORG ISOLATED GASTROCNEMIUS TIGHTNESS VOLUME 84-A · NUMBER 6 · JUNE 2002

contracture, whereas approximately one-quarter of the asymp- Nirmal Tejwani, MD tomatic individuals had gastrocnemius contracture and one-sixth Department of Orthopaedics, New York University School of Medicine, demonstrated a combined contracture. Clinical examination 101 Commodore Terrace, Edgewater, NJ 07020 was demonstrated to be fairly reliable in identifying muscle Robert Price, MSME tightness. Joseph Cziernecki, MD Although our data are clearly preliminary, such find- Department of Rehabilitation Medicine, University of Washington, ings may have implications for preventative care and should RR&D Center, Mailstop 151, Veteran’s Affairs Puget Sound Health Care heighten awareness of the existence and potential long-term System, 1660 South Columbian Way, Seattle, WA 98108 effects of tightness of the muscles, particularly the gastroc- nemius muscle, in patients without neurological impairment. Sigvard T. Hansen Jr., MD Bruce J. Sangeorzan, MD We suspect that this pathological entity plays a vital role in Department of Orthopaedics, University of Washington, 325 Ninth Ave- chronic mechanical breakdown or inflammation of both the nue, Seattle, WA 98104-2499 foot and ankle.  In support of their research or preparation of this manuscript, one or more of the authors received a grant from the Department of Veterans Affairs, Seattle Veterans Affairs Medical Center, Seattle, Washington. The Christopher W. DiGiovanni, MD equipment was developed with use of a Department of Veterans Affairs Department of Orthopaedics, Brown University School of Medicine, Center Grant, although no revenues were received. None of the authors University Orthopedics Incorporated, 1287 North Main Street, Provi- received payments or other benefits or a commitment or agreement to dence, RI 02904 provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research Roderick Kuo, MD fund, foundation, educational institution, or other charitable or non- P.O. Box 913, Strathfield, Sydney, NSW, Australia 2135 profit organization with which the authors are affiliated or associated.

References

1. Banks HH, Green WT. The correction of equinus deformity in cerebral palsy. 16. Strayer LM Jr. Recession of the gastrocnemius. An operation to relieve spas- J Bone Joint Surg Am. 1958;40:1359-79. tic contracture of the calf muscles. J Bone Joint Surg Am. 1950;32:671-6. 2. Damron TA, Greenwald TA, Breed AL. Chronologic outcome of surgical ten- 17. Silfverskiold N. Reduction of the uncrossed two- muscles of the leg to doachilles lengthening and natural history of gastroc-soleus contracture in one-joint muscles in spastic conditions. Acta Chir Scand. 1924;56:315-30. cerebral palsy. A two-part study. Clin Orthop. 1994;301:249-55. 18. Vulpius O, Stoffel A. Orthopadische Operationslehre. Stuttgart: Ferdinand 3. Nicoladoni C. The classic. On the treatment of pes equinus paralyticus. Clin Enke; 1913. Tenotomie der end schnen der mm. gastrocnemius el soleus Orthop. 1978;135:2-3. mittels rutschenlassens nach vulpius; p 29-31. 4. Delpech J. Chirurgie clinique de Montpellier; où, observations et reflexions 19. Craig JJ, van Vuren J. The importance of gastrocnemius recession in the tirées des travaux de chirurgie clinique de cette école. Paris: Gabon; 1823-8. correction of equinus deformity in cerebral palsy. J Bone Joint Surg Br. p 147-231. 1976;58:84-7. 5. Baker LD. A rational approach to the surgical needs of the cerebral palsy pa- 20. Sharrard WJ, Bernstein S. Equinus deformity in cerebral palsy. A comparison tient. J Bone Joint Surg Am. 1956;38:313-23. between elongation of the tendo and gastrocnemius recession. 6. Beckerman H, Becher J, Lankhorst GJ, Verbeek AL. Walking ability of stroke JBone Joint Surg Br. 1972;54:272-6. patients: efficacy of tibial blocking and a polypropylene ankle-foot 21. Smetana V, Schejbalova A. [The Strayer surgical technic as the basic opera- orthosis. Arch Phys Med Rehabil. 1996;77:1144-51. tion for treatment of pes equinus in cerebral palsy]. Acta Chir Orthop Trauma- 7. Camacho FJ, Isunza A, Coutino B. Comparison of tendo-Achilles lengthening tol Cech. 1993;60:218-24. Czech. alone and combined with neurectomy of the gastrocnemius muscle in the 22. Brodersen A, Pedersen B, Reimers J. [Foot deformities and relation to the treatment of equinus deformity of the foot associated with clonus in children length of leg muscles in Danish children aged 3-17 years]. Ugeskr Laeger. with cerebral palsy. Orthopedics. 1996;19:319-22. 1993;155:3914-6. Danish. 8. Conine TA, Sullivan T, Mackie T, Goodman M. Effect of serial casting for the 23. Downey MS, Banks AS. Gastrocnemius recession in the treatment of non- prevention of equinus in patients with acute head injury. Arch Phys Med spastic ankle equinus. A retrospective study. J Am Podiatr Med Assoc. Rehabil. 1990;71:310-2. 1989;79:159-74. 9. Corry IS, Cosgrove AP, Duffy CM, McNeill S, Taylor TC, Graham HK. Botuli- 24. Grady JF, Saxena A. Effects of stretching the gastrocnemius muscle. J Foot num toxin A compared with stretching casts in the treatment of spastic equi- Surg. 1991;30:465-9. nus: a randomised prospective trial. J Pediatr Orthop. 1998;18:304-11. 25. Hillstrom HJ, Perlberg G, Siegler S, Sanner WH, Hice GA, Downey M, Stien- 10. Feehery RV Jr. Surgery of the achilles tendon and posterior muscle group. stra J, Acello A, Neary MT, Kugler F. Objective identification of ankle equinus Clin Podiatr Med Surg. 1992;9:781-811. deformity and resulting contracture. J Am Podiatr Med Assoc. 1991;81:519-24. 11. Graham HK, Fixsen JA. Lengthening of the calcaneal tendon in spastic hemi- 26. Root ML, Orien WP, Weed JH. Normal and abnormal function of the foot: clini- plegia by the White slide technique. A long-term review. J Bone Joint Surg Br. cal biomechanics. Vol 2. Los Angeles: Clinical Biomechanics Corp., 1977. 1988;70:472-5. p 174. 12. Javors JR, Klaaren HE. The Vulpius procedure for correction of equinus defor- 27. Sgarlato TE, Morgan J, Shane HS, Frenkenberg A. Tendo achillis lengthening mity in cerebral palsy. J Pediatr Orthop. 1987;7:191-3. and its effect on foot disorders. J Am Podiatry Assoc. 1975;65:849-71. 13. McCall RE, Frederick HA, McCluskey GM, Riordan DC. The Bridle proce- 28. Subotnick SI. Equinus deformity as it affects the forefoot. J Am Podiatry dure: a new treatment for equinus and equinovarus deformities in children. Assoc. 1971;61:423-7. JPediatr Orthop. 1991;11:83-9. 29. Harris RI, Beath T. Hypermobile flat-foot with short tendo Achillis. J Bone 14. Olney BW, Williams PF, Menelaus MB. Treatment of spastic equinus by apo- Joint Surg Am. 1948;30:116-38. neurosis lengthening. J Pediatr Orthop. 1988;8:422-5. 30. Hibbs RA. Muscle bound feet. New York Med J. 1914;100:797. 15. Sindou M, Mertens P. Selective neurotomy of the for treatment of the spastic foot. Neurosurgery. 1988;23:738-44. 31. Mann RA, Coughlin MJ. Hallux valgus—etiology, anatomy, treatment and sur-

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014 

THE JOURNAL OF BONE & JOINT SURGERY · JBJS.ORG ISOLATED GASTROCNEMIUS TIGHTNESS VOLUME 84-A · NUMBER 6 · JUNE 2002

gical considerations. Clin Orthop. 1981;157:31-41. 42. Davis WL Jr, Singerman R, Labropoulos PA, Victoroff B. Effect of ankle and knee position on tension in the Achilles tendon. Foot Ankle Int. 1999;20:126-31. 32. Toolan BC, Sangeorzan BJ, Hansen ST Jr. Complex reconstruction for the treat- ment of dorsolateral peritalar subluxation of the foot. Early results after distrac- 43. Johnson KA, Strom DE. Tibialis posterior tendon dysfunction. Clin Orthop. tion arthrodesis of the calcaneocuboid joint in conjunction with stabilization of, 1989;239:196-206. and transfer of the flexor digitorum longus tendon to, the midfoot to treat ac- 44. Armstrong DG, Stacpoole-Shea S, Nguyen H, Harkless LB. Lengthening of quired pes planovalgus in adults. J Bone Joint Surg Am. 1999;81:1545-60. the Achilles tendon in diabetic patients who are at high risk for ulceration of 33. Arkkila PE, Kantola IM, Viikari JS. Limited joint mobility in type I diabetic the foot. J Bone Joint Surg Am. 1999;81:535-8. patients: correlation to other diabetic complications. J Intern Med. 1994; 45. Morton DJ. The human foot: its evolution, physiology, and functional disorders. 236:215-23. New York: Columbia University; 1935. 34. Green DR, Ruch JA, McGlamry ED. Correction of equinus-related forefoot 46. Fraser RK, Menelaus MB, Williams PF, Cole WG. The Miller procedure for deformities: a case report. J Am Podiatry Assoc. 1976;66:768-80. mobile flat feet. J Bone Joint Surg Br. 1995;77:396-9. 35. Hill RS. Ankle equinus. Prevalence and linkage to common foot pathology. 47. Lin SS, Lee TH, Wapner KL. Plantar forefoot ulceration with equinus deformity J Am Podiatr Med Assoc. 1995;85:295-300. of the ankle in diabetic patients: the effect of tendo-Achilles lengthening and 36. Katoulis EC, Boulton AJ, Raptis SA. The role of diabetic neuropathy and high total contact casting. Orthopedics. 1996;19:465-75. plantar pressures in the pathogenesis of foot ulceration. Horm Metab Res. 48. Sinacore DR. Healing times of diabetic ulcers in the presence of fixed defor- 1996;28:159-64. mities of the foot using total contact casting. Foot Ankle Int. 1998;19:613-8. 37. Mueller MJ, Diamond JE, Delitto A, Sinacore DR. Insensitivity, limited joint 49. Barry DC, Sabacinski KA, Habershaw GM, Giurini JM, Chrzan JS. Tendo mobility, and plantar ulcers in patients with diabetes mellitus. Phys Ther. Achillis procedures for chronic ulcerations in diabetic patients with trans- 1989;60:453-62. metatarsal amputations. J Am Podiatr Med Assoc. 1993;83:96-100. 38. Simmons RW, Richardson C, Deutsch K. Limited joint mobility of the ankle in 50. Frykberg RG, Lavery LA, Pham H, Harvey C, Harkless L, Veves A. Role of diabetic patients with cutaneous sensory deficit. Diabetes Res Clin Pract. neuropathy and high foot pressures in diabetic foot ulceration. Diabetes 1997;37:137-43. Care. 1998;21:1714-9. 39. Lehmann JF, Price R, deLateur BJ, Hinderer S, Traynor C. Spasticity: quanti- 51. Ward ED, Phillips RD, Patterson PE, Werkhoven GJ. 1998 William J. Stickel tative measurements as a basis for assessing effectiveness of therapeutic Gold Award. The effects of extrinsic muscle forces on the forefoot-to-rearfoot intervention. Arch Phys Med Rehabil. 1989;70:6-15. loading relationship in vitro. and Achilles tendon. J Am Podiatr Med 40. Price R, Bjornson KF, Lehmann JF, McLaughlin JF, Hays RM. Quantitative Assoc. 1998;88:471-82. measurement of spasticity in children with cerebral palsy. Dev Med Child Neu- 52. Fulp MJ, McGlamry ED. Gastrocnemius tendon recession. Tongue in groove rol. 1991;33:585-95. procedure to lengthen gastrocnemius tendon. J Am Podiatry Assoc. 1974; 41. Weaver K, Price R, Cziernecki J, Sangeorzan B. Design and validation of an 64:163-71. instrument package designed to increase the reliability of ankle range of mo- 53. Whittle M. Gait analysis: an introduction. 2nd ed. Oxford: Butterworth- tion measurements. J Rehabil Res Dev. 2001;38:471-5. Heinemann; 1996. p 58.

Downloaded From: http://jbjs.org/ by a UNIVERSITY OF WASHINGTON User on 03/05/2014